69%
reported individual use
In 8am’s 2026 survey research, reported use of general-purpose generative AI among legal professionals more than doubled from the prior year.
AI Adoption and Readiness Assessment for Law Firms
Fixed-fee assessment
Turn informal AI use into a governed advantage for your firm.
Identify how AI is already being used, protect sensitive information, qualify workflows worth pursuing, establish verification and ownership, and define a bounded pilot before the firm commits to a rollout or another platform.
The first conversation is a practical fit discussion—not a commitment to purchase software, prohibit AI, begin implementation, or change providers.
Questions this assessment answers
- What AI tools and embedded capabilities are people already using—formally or informally?
- What client, matter, firm, or public information is entering those tools?
- Which capabilities already exist in platforms and licenses the firm trusts and pays for?
- Which workflows justify AI based on value, risk, readiness, and measurable outcomes?
- How will the firm verify authority, demonstrate professional judgment, and assign ownership?
Leadership readiness checklist
Five questions your firm should be able to answer.
If one or more answers are unclear, the firm has a visibility or governance gap—not simply a need for another AI product.
- 1
Can leadership identify which AI tools, accounts, and embedded features attorneys and staff currently use?
- 2
Can the firm explain what client, matter, personnel, firm, or public information enters each environment?
- 3
Has the firm defined approved accounts and configurations—and when a use is prohibited, controlled, or routine with baseline safeguards?
- 4
Can reviewers demonstrate authoritative-source checks, citation and quotation verification, and actual professional judgment?
- 5
Is there an accountable owner for policy, training, exceptions, incidents, pilot measures, and periodic review?
Who this is for
Built for the leadership decisions smaller firms are facing now.
The assessment is designed for managing partners, firm owners, and practice leaders who need a dependable operating view—not another generalized AI presentation.
- Leadership lacks a firm-wide view of informal AI use, accounts, information, and workarounds.
- The firm is choosing an AI platform, defining policy boundaries, or deciding whether a pilot is justified.
- A policy or training effort exists, but verification, enforcement, exceptions, ownership, or measurement remain unclear.
Illustrative sample
See the form of the decisions—not just a list of deliverables.
This fictional two-page excerpt shows how assessment evidence can be organized into a leadership brief and a bounded 90-day roadmap.
The full engagement adds validated evidence, firm-specific use and vendor inventories, workflow qualification, an authority-review register, and findings reviewed with leadership.
View the sample executive summary (PDF, 2 pages)Illustrative sample only—not client work, legal advice, an ethics opinion, certification, or compliance attestation.
Adoption evidence
Individual use is moving faster than firm-wide coordination.
The opportunity is no longer simply to experiment. Leadership needs a coordination layer that connects technology, information, governance, training, verification, and measurable workflow value.
54%
reported no planned AI training
Access without practical training leaves people to make inconsistent decisions about information, verification, and appropriate use.
43% / 9%
policy and enforcement remain thin
Forty-three percent reported no planned formal AI policy, while only nine percent reported a policy that was actively enforced.
These findings come from vendor-sponsored 8am research and describe its surveyed respondents, not every lawyer or law firm. Read the 8am 2026 legal-industry report.
What uncertainty can cost
Ungoverned adoption creates exposure while governed adoption can create measurable capacity.
The cost is not limited to a bad answer. Informal use can alter workflows, information handling, review habits, accountability, licensing, and client expectations without leadership seeing the change.
Invisible workflow change
People may already use AI for research, summaries, drafts, analysis, or administration while the firm lacks an inventory, boundaries, or a dependable way to learn from actual use.
Confidentiality exposure
Risk depends on the information, task, account configuration, contract, retention and training terms, access controls, and where data is processed—not simply the product name.
Plausible-but-wrong work
Fluent output can hide fabricated authority, inaccurate quotations, misstated holdings, omitted context, or unsupported conclusions that survive a superficial review.
Diffused accountability
Vendor selection, IT ownership, or a lawyer being somewhere in the workflow does not show who verified the work or exercised judgment.
Redundant licensing
Standalone tools can duplicate capabilities already present in approved practice, research, productivity, or document platforms.
Capacity without proof
Claims of saved time or reduced effort are difficult to defend without a baseline, representative work, quality measures, exception handling, and explicit stop conditions. A pilot should show whether AI reduces total effort, preserves quality, and redirects attorney or support-staff time to higher-value work.
A policy is one control, not the operating system. The assessment connects approved use, technology configuration, authoritative sources, training, review, ownership, monitoring, and pilot evidence.
South Carolina context · Last reviewed August 31, 2026
Use current South Carolina signals without overstating them.
The legal landscape is moving quickly. The assessment records current sources and turns them into technology and workflow questions while the firm’s counsel determines legal and ethical applicability.
Judicial Branch interim policy
The March 25, 2025 interim policy applies directly to South Carolina Judicial Branch employees and volunteers. It requires approved tools, human oversight, independent verification, and restrictions around confidential or privileged information.
Private-practice distinction
The interim policy does not directly bind private practitioners, but it reminds lawyers appearing before South Carolina courts to exercise caution, verify accuracy, and preserve confidentiality under existing professional obligations.
Technology competence
South Carolina Rule of Professional Conduct 1.1 includes the duty to keep abreast of changes in law and practice, including the benefits and risks associated with relevant technology.
Filing disclosure
As of this review, no statewide South Carolina state-court rule requiring attorneys to disclose AI use in filings was identified. Matter-specific, judge-specific, federal, and later-issued requirements still require counsel’s review.
Carolina Technology Pros provides technology-governance and workflow guidance—not legal advice, an ethics opinion, filing certification, or compliance attestation. Counsel determines which duties, rules, orders, and client commitments apply.
Risk and accountability
Design controls around the work—not around confidence in a brand.
General-purpose and legal-specific tools can both be useful, and neither category removes the firm’s responsibility to qualify the environment, verify the work, and respond appropriately when something goes wrong.
Unreliable authority or analysis
AI can invent cases, create quotations that do not exist, misstate a real holding, omit controlling context, or produce a confident conclusion unsupported by the cited source.
Information in the wrong environment
Client, matter, personnel, financial, or firm information may be exposed when the task, account type, contract, configuration, access, retention, or vendor handling has not been approved.
Review without demonstrable judgment
A human-in-the-loop statement is weak if the firm cannot show the authoritative sources checked, the reviewer’s responsibility, the exceptions resolved, and the final decision made.
Errors aggravated by the response
A mistake can become more consequential when people conceal tool use, provide inaccurate explanations, fail to correct the record promptly, or cannot reconstruct what happened.
Selected enforcement examples
Verification, accountability, and the response when errors surface all matter.
Mata v. Avianca (S.D.N.Y. 2023)
The court imposed a $5,000 penalty after attorneys submitted nonexistent decisions and then stood by false material through inadequate and misleading follow-up. The opinion addressed both the filing and the response after doubts were raised.
Review the Mata sanctions discussion in U.S. Courts materials.Fletcher v. Experian (Fifth Circuit 2026)
The court imposed a $2,500 sanction after a brief contained 16 fabricated quotations. It expressly noted that greater candor about the role of generative AI likely would have produced a lesser sanction.
Read the Fifth Circuit opinion (PDF).Whiting v. Welles-Bowen Realty (Sixth Circuit 2026)
In a matter involving fabricated quotations and broader litigation conduct, the court imposed $15,000 on each attorney, fees and double costs, and referrals. The result should not be reduced to a single tool error.
Read the Sixth Circuit opinion (PDF).Fourth Circuit public admonishment (2026)
The court publicly admonished counsel after briefing cited nonexistent cases that appeared potentially AI-generated, emphasizing that existing duties of competence, candor, and verification already address the conduct.
Read the Fourth Circuit opinion (PDF).The examples involve different facts, conduct, procedures, and remedies. They do not establish that every error was exclusively caused by AI or that every matter leads to the same sanction.
Proposed operational framework
Match permission and oversight to the real risk of the use.
The assessment classifies uses by task, data, audience, consequence, account configuration, authoritative sources, and required professional judgment—not by product brand alone.
Prohibited
- Representative uses
- Restricted or confidential matter information in an unapproved consumer environment; unverified authority in a filing; autonomous decisions about client rights, strategy, or outcomes.
- Required treatment
- Do not permit. Define the boundary, block or remove access where practical, give people a safe escalation path, and address discovered use promptly.
Controlled
- Representative uses
- Legal research, matter summarization, document review, discovery analysis, first drafts, and client-facing material where error or disclosure could be consequential.
- Required treatment
- Use an approved environment, explicit information boundaries, authoritative-source verification, accountable attorney review, documented exceptions, and appropriate monitoring.
Routine with baseline controls
- Representative uses
- Scheduling, generic administration, nonconfidential brainstorming, and generic marketing or internal communications with limited consequence.
- Required treatment
- Use an approved account and normal review. Preserve confidentiality, advertising and brand standards, access controls, and the prohibition on unsupported legal claims.
This is the assessment’s proposed operational framework. It is informed by guidance emphasizing competence, security, supervision, confidentiality, and independent professional judgment; it is not attributed to or presented as a North Carolina State Bar tiering system. Read North Carolina 2024 Formal Ethics Opinion 1.
Dated reference notes
Sources leadership and counsel can review.
These links support the dated context on this page. They are reference points, not a substitute for counsel’s review of current, court-specific, client-specific, carrier-specific, or matter-specific requirements.
Last reviewed August 31, 2026
- ABA Formal Opinion 512 (PDF)
Addresses lawyers’ use of generative AI under duties including competence, confidentiality, communication, candor, supervision, and reasonable fees.
- Stanford RegLab legal-research benchmark
The reported error rates describe the tested product versions and query sets from the 2024 study—not current 2026 product performance.
- Damien Charlotin AI hallucination case tracker
Tracks decisions addressing established or suspected hallucinated material; the collection is broader than attorney-sanction decisions and its count changes frequently.
- South Carolina Bar ethics advisory opinions
A current index for counsel to check as South Carolina guidance develops.
When to take a closer look
Warning signs worth investigating before they become urgent.
- Lawyers and staff use AI without a firm-wide inventory or shared boundaries
- A policy exists on paper but training, enforcement, exceptions, and monitoring are undefined
- Confidential or matter information may enter tools without an approved configuration or vendor review
- AI-assisted research or drafting lacks an authoritative-source verification protocol
- New licenses are considered before existing platform capabilities and workflow needs are understood
- The firm cannot describe who owns AI decisions, incidents, training, or pilot results
What we examine
The operating conditions behind the decisions.
Current formal and informal use
Who is using what, for which tasks, with which information, under what account, and with what known workarounds, adoption patterns, or avoided uses.
Tool, license, and vendor fit
Existing platform capabilities, overlapping licenses, account configurations, contracts, data handling, supportability, and qualification criteria for additional vendors.
Information handling
Client and matter confidentiality, privilege-sensitive workflows, data classification, access, retention, training use, sharing, deletion, and incident escalation.
Workflow and authoritative sources
The actual task, source of truth, inputs, outputs, handoffs, exceptions, audiences, and consequences when an answer is incomplete or wrong.
Verification and professional review
Citation and quotation checks, source comparison, review evidence, decision ownership, candor, correction, and the point at which professional judgment must be exercised.
Governance, training, and measurement
Approved-use decisions, policy operations, role-based training, enforcement, monitoring, incident response, pilot baselines, quality measures, value, and stop conditions.
Assessment fit
Choose the assessment that matches the decision.
AI governance and workflow modernization overlap, but they answer different leadership questions.
AI Adoption and Readiness Assessment
Use this assessment to govern firm-wide AI use, information handling, verification, ownership, platform decisions, and bounded pilots.
Continue reviewing AI readinessBusiness Process Modernization Assessment
Use this assessment to redesign one high-friction workflow before automating it or selecting implementation technology.
Review process modernizationWhat you receive
A practical, decision-ready package.
Leadership receives operational decision tools—not a generic policy template, software recommendation, or promise that AI is risk-free.
AI-use inventory
A practical record of known formal and informal uses, tools, accounts, information classes, users, owners, and material gaps.
Use-case portfolio
Candidate workflows prioritized by value, readiness, risk, professional judgment, supportability, and measurement potential.
Vendor qualification criteria
Decision criteria for configuration, contract, information handling, security, integration, support, exit, and accountable ownership.
Approved-use matrix
Clear permitted, controlled, and prohibited uses tied to information, audience, environment, review, and escalation requirements.
Verification protocol
Authoritative-source checks, citation and quotation validation, reviewer responsibilities, exception handling, correction, and evidence expectations.
Training and enforcement plan
Role-based learning, acknowledgments, reinforcement, exception handling, monitoring, response, and accountable ownership.
Bounded pilot definition
Representative work, controlled inputs, approved users, baselines, quality and value measures, review steps, stop conditions, and decision gate.
Authority-review register
A maintained list of relevant ethics opinions, court rules, standing orders, client commitments, insurer conditions, and sources assigned for legal review.
Adoption roadmap
A sequenced path for governance, configuration, training, pilots, measurement, platform decisions, ownership, and periodic review.
A responsible outcome may be to approve, control, prohibit, pilot, optimize an existing capability, defer a purchase, or stop a use. The workflow evidence and the firm’s legal review determine the path.
How the assessment works
Evidence first. Decisions before implementation.
The engagement is defined before paid work begins so participants, evidence, confidentiality boundaries, legal-review responsibilities, decisions, timing, and deliverables are understood.
- 1
Fit and scope
Define the leadership decision, practices and workflows in scope, participants, boundaries, legal-review needs, evidence, deliverables, timing, and investment.
- 2
Evidence collection
Interview leaders and representative users; inventory tools, accounts, licenses, configurations, policies, training, workflows, sources, incidents, and available measures.
- 3
Analysis and design
Qualify use cases, information environments, vendors, verification controls, ownership, training, enforcement, measures, and a bounded pilot.
- 4
Findings review
Review the inventory, risk decisions, evidence gaps, roadmap, authority-review register, pilot gate, and the choices leadership and counsel still need to make.
A first-class paid assessment
A professional decision package that stands on its own.
The first conversation is a brief no-cost fit and scope discussion. Evidence collection, analysis, scoring, findings, and roadmapping begin only after a written scope and fixed fee are approved.
The completed assessment organizes the material findings, risks, responsibilities, decisions, and sequenced roadmap so leadership can act with an internal team, Carolina Technology Pros, an existing provider, or another qualified provider. Implementation is optional, priced separately, and the assessment fee is not an implementation deposit or automatic credit.
Clear boundaries
An assessment—not legal advice, certification, or an automatic rollout.
Carolina Technology Pros organizes technology, information, vendor, workflow, governance, and measurement decisions. The firm and its counsel retain responsibility for professional obligations and legal conclusions.
- The assessment does not provide a legal opinion, ethics opinion, filing certification, compliance attestation, or guarantee that a use is permitted.
- Firm-specific policy language, client disclosures, privilege decisions, court obligations, and malpractice-insurance implications remain subject to counsel and carrier review.
- Named tools are not treated as uniformly safe or unsafe; account configuration, contract, controls, information, task, and professional review all matter.
- Implementation, configuration, migration, training delivery, monitoring, managed services, and software purchasing require separate authorization after findings are reviewed.
How Carolina Technology Pros approaches the work
Tool-neutral guidance that starts with the firm’s actual operating environment.
The assessment begins with the work, information, authoritative sources, people, current platforms, and leadership decisions—not a predetermined software sale.
- Appropriate capabilities in existing trusted platforms are evaluated before new purchases are recommended.
- The firm’s current IT provider, practice-management consultant, research provider, internal team, counsel, or insurer can participate where their evidence and responsibilities matter.
- Recommendations remain usable by the firm and its chosen providers; implementation is optional and separately authorized.
- Evidence gaps and unresolved legal questions stay visible instead of being converted into false certainty.
Go deeper
Related guidance for this decision.
Use these practical resources to prepare questions, compare options, and decide what evidence should come next.
How to Choose an AI Platform for Your Business
Compare ChatGPT, Claude, Copilot, and Gemini by business fit, governance, cost, and adoption before choosing a platform and planning a controlled pilot.
Read resourceThe AI Model Is Not Your Moat: Build the Learning Loop Your Business Owns
Assess whether your AI creates a learning loop—and how business-specific systems, employee training, and expert guidance can build an advantage.
Read resourceWhat to Put in Place Before Adopting AI Tools
Evaluate use cases, information sensitivity, human review, governance, and proof-of-value before broad AI adoption.
Read resourceAn AI Workflow Is Not a Legal Research Authority
Separate AI drafting from controlling law, maintained legal guidance, firm precedents, matter records, and accountable attorney verification clearly.
Read resourceClaude, Copilot, or Legal AI? Start With the Workflow
Compare legal AI choices through authoritative content, governance, integration, support, portability, and verified reviewer effort before selection.
Read resourceCan Our Law Firm Use Generative AI Safely? A Practical Governance Checklist
Build a law-firm AI operating boundary around confidentiality, verification, supervision, client communication, fees, approved tools, and accountable review.
Read resourceQuestions business leaders ask
What to expect before you begin.
Is this legal advice or an AI compliance certification?
No. Carolina Technology Pros provides technology-governance and workflow guidance. The assessment is not legal advice, an ethics opinion, filing certification, or compliance attestation. Your firm’s counsel determines applicability of rules, duties, orders, contracts, and client commitments.
Does responsible AI adoption mean the firm must prohibit AI?
No. Some uses may be prohibited, some may require strict controls, and others may be routine with baseline safeguards. The decision follows the task, information, audience, consequence, configuration, authoritative sources, and required judgment.
Can confidential client information be used with AI?
That is not a product-name question and this assessment does not make the legal determination. The firm must consider confidentiality and privilege obligations alongside the specific account, contract, retention and training terms, access, security, processing, task, and approved review. Counsel should decide the firm-specific rule.
Will the assessment choose an AI platform?
It can produce vendor qualification criteria and a fit recommendation when evidence supports one. Existing approved platform capabilities are examined first, and no named platform is assumed to be uniformly safe, unsafe, approved, or endorsed.
Does the assessment produce an AI policy?
It produces the evidence, approved-use matrix, ownership model, verification protocol, training and enforcement plan, and authority-review register needed to support a workable policy. Firm-specific legal language should be finalized by counsel.
Can our current IT provider participate?
Yes. Current providers can contribute inventories, configurations, contracts, security and support evidence, and operating context. The roadmap can be used by the firm, its existing providers, Carolina Technology Pros, or another qualified team.
Is implementation included?
No. The assessment defines decisions, controls, ownership, measures, and a roadmap. Configuration, software procurement, pilot execution, training delivery, monitoring, and ongoing services are separately authorized and scoped.
How are timing and investment determined?
After the fit conversation, a written scope confirms the firm areas, participants, evidence, legal-review responsibilities, deliverables, exclusions, timing, and investment before paid work begins.
A practical first conversation
Start with the AI decision your firm needs to make.
Share the workflow, policy gap, platform question, client concern, or pilot decision leadership needs to resolve. The first conversation will determine fit—without committing the firm to a purchase or rollout.
If you want implementation help
Use the roadmap with the right delivery team.
Your internal team, current provider, Carolina Technology Pros, or another qualified partner can implement the work. Any delivery engagement is optional and scoped separately.